Loading examples/hyperparam_opt/gaussian_hyperparam_opt.py +3 −8 Original line number Diff line number Diff line Loading @@ -11,13 +11,8 @@ train, valid, test= delaney_datasets # Fit models regression_metric = dc.metrics.Metric(dc.metrics.pearson_r2_score) # TODO(rbharath): I don't like this awkward string/class divide. Maybe clean up? optimizer = dc.hyper.GaussianProcessHyperparamOpt('tf_regression') best_hyper_params, best_performance = optimizer.hyperparam_search( dc.molnet.preset_hyper_parameters.hps['tf_regression'], train, valid, transformers, [regression_metric] ) dc.molnet.preset_hyper_parameters.hps['tf_regression'], train, valid, transformers, [regression_metric]) Loading
examples/hyperparam_opt/gaussian_hyperparam_opt.py +3 −8 Original line number Diff line number Diff line Loading @@ -11,13 +11,8 @@ train, valid, test= delaney_datasets # Fit models regression_metric = dc.metrics.Metric(dc.metrics.pearson_r2_score) # TODO(rbharath): I don't like this awkward string/class divide. Maybe clean up? optimizer = dc.hyper.GaussianProcessHyperparamOpt('tf_regression') best_hyper_params, best_performance = optimizer.hyperparam_search( dc.molnet.preset_hyper_parameters.hps['tf_regression'], train, valid, transformers, [regression_metric] ) dc.molnet.preset_hyper_parameters.hps['tf_regression'], train, valid, transformers, [regression_metric])
